{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:C5LRE5UYJIBUXP7SX5OY43SMUF","short_pith_number":"pith:C5LRE5UY","schema_version":"1.0","canonical_sha256":"17571276984a034bbff2bf5d8e6e4ca17b7602c63801dd3c9a545f3472dc0907","source":{"kind":"arxiv","id":"2503.01034","version":1},"attestation_state":"computed","paper":{"title":"Data Unlearning in Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kenan Hasanaliyev, Manav Shah, Silas Alberti, Stefano Ermon","submitted_at":"2025-03-02T21:36:04Z","abstract_excerpt":"Recent work has shown that diffusion models memorize and reproduce training data examples. At the same time, large copyright lawsuits and legislation such as GDPR have highlighted the need for erasing datapoints from diffusion models. However, retraining from scratch is often too expensive. This motivates the setting of data unlearning, i.e., the study of efficient techniques for unlearning specific datapoints from the training set. Existing concept unlearning techniques require an anchor prompt/class/distribution to guide unlearning, which is not available in the data unlearning setting. Gene"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2503.01034","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-02T21:36:04Z","cross_cats_sorted":[],"title_canon_sha256":"e1308f18c7f5f3b34f75cdf94aaef8ccccba486f864a8028be874d3a2fbd3cda","abstract_canon_sha256":"7925f8ec90e8923e8947d83322acd641b6e70d3cb1184493607d6d8f8a453583"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:22:34.870132Z","signature_b64":"Gzlfc31LLLEjhYx5YDTZcyJNjov54dnDqC4gh2+6PdFrU7IXRwVT+VWUANJlxoWPWzlusVg6xWlnqLnRedadBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"17571276984a034bbff2bf5d8e6e4ca17b7602c63801dd3c9a545f3472dc0907","last_reissued_at":"2026-07-05T10:22:34.869646Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:22:34.869646Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Data Unlearning in Diffusion Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kenan Hasanaliyev, Manav Shah, Silas Alberti, Stefano Ermon","submitted_at":"2025-03-02T21:36:04Z","abstract_excerpt":"Recent work has shown that diffusion models memorize and reproduce training data examples. At the same time, large copyright lawsuits and legislation such as GDPR have highlighted the need for erasing datapoints from diffusion models. However, retraining from scratch is often too expensive. This motivates the setting of data unlearning, i.e., the study of efficient techniques for unlearning specific datapoints from the training set. Existing concept unlearning techniques require an anchor prompt/class/distribution to guide unlearning, which is not available in the data unlearning setting. Gene"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.01034","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2503.01034/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2503.01034","created_at":"2026-07-05T10:22:34.869707+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.01034v1","created_at":"2026-07-05T10:22:34.869707+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.01034","created_at":"2026-07-05T10:22:34.869707+00:00"},{"alias_kind":"pith_short_12","alias_value":"C5LRE5UYJIBU","created_at":"2026-07-05T10:22:34.869707+00:00"},{"alias_kind":"pith_short_16","alias_value":"C5LRE5UYJIBUXP7S","created_at":"2026-07-05T10:22:34.869707+00:00"},{"alias_kind":"pith_short_8","alias_value":"C5LRE5UY","created_at":"2026-07-05T10:22:34.869707+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.30825","citing_title":"Unlearning in Diffusion Models: A Unified Framework with KL Divergence and Likelihood Constraints","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2405.07406","citing_title":"Machine Unlearning: A Comprehensive Survey","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07962","citing_title":"Is your algorithm unlearning or untraining?","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF","json":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF.json","graph_json":"https://pith.science/api/pith-number/C5LRE5UYJIBUXP7SX5OY43SMUF/graph.json","events_json":"https://pith.science/api/pith-number/C5LRE5UYJIBUXP7SX5OY43SMUF/events.json","paper":"https://pith.science/paper/C5LRE5UY"},"agent_actions":{"view_html":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF","download_json":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF.json","view_paper":"https://pith.science/paper/C5LRE5UY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.01034&json=true","fetch_graph":"https://pith.science/api/pith-number/C5LRE5UYJIBUXP7SX5OY43SMUF/graph.json","fetch_events":"https://pith.science/api/pith-number/C5LRE5UYJIBUXP7SX5OY43SMUF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF/action/storage_attestation","attest_author":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF/action/author_attestation","sign_citation":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF/action/citation_signature","submit_replication":"https://pith.science/pith/C5LRE5UYJIBUXP7SX5OY43SMUF/action/replication_record"}},"created_at":"2026-07-05T10:22:34.869707+00:00","updated_at":"2026-07-05T10:22:34.869707+00:00"}